Decision Support for Natural Ventilation

of Nonresidential Buildings
Ying Zhao1 and James R. Jones2
Abstract: This paper describes a decision-support framework assisting the design of nonresidential buildings with natural ventilation.
The framework is composed of decision modules with input, analysis algorithms and output of natural ventilation design. The framework
covers ventilation with natural driving force and mechanical-assisted ventilation. The proposed framework has two major assessment
levels: feasibility assessment and comparison of alternative natural ventilation approaches. The feasibility assessment modules assess the
potential of the site with the design proposition for natural ventilation in terms of wind, temperature, humidity, noise, and pollution
conditions. All of the possible natural ventilation approaches and system designs are assessed by first applying constraint functions to each
of the alternatives. Then the comparison of alternative approaches to natural ventilation continues by assessing the critical performance
mandates that include energy savings, thermal comfort, acoustic control, indoor air quality, and cost. Approaches are finally ranked based
on their performance.
DOI: 10.1061/ASCE1076-0431200713:295
CE Database subject headings: Ventilation; Design; Decision support systems; Decision making; Buildings, nonresidential.

Introduction
As a design strategy, when properly understood and applied, natural ventilation can significantly reduce building energy consumption while ensuring both good indoor air quality IAQ and
thermal comfort conditions. Natural ventilation as a design and
operating strategy may be effective in many climate zones in the
United States Chen 2004; Jones and West 2001. Unfortunately,
natural ventilation is currently underutilized in nonresidential
buildings. The paper represents a first step toward developing a
decision-support framework and therefore reader feedback is encouraged through the writer contact given at the end of the paper.
When and how to integrate natural ventilation approaches into
nonresidential buildings are not well known. To be most effective,
the design of naturally ventilated buildings must begin early in
the design process such as choosing and analyzing the site leading
to the assessment of the performance of alternative approaches.
The decisions related to this process must take into account both
technical and architectural goals, as well as understanding of performance issues such as energy savings, thermal comfort, sound
transmission, security, indoor air quality, and cost.
This paper identifies critical issues and proposes assessment
approaches with the inputs, analysis algorithms and outputs for a
decision-support structure for natural ventilation of nonresidential
1

Ph.D. Candidate, Dept. of Architecture, Virginia Tech, 300 Cowgill

Hall, Blacksburg, VA 24061 corresponding author. E-mail: yizhao1@
vt.edu
2
Associate Professor, Dept. of Architecture, Virginia Tech, 300B
Cowgill Hall, Blacksburg, VA 24061.
Note. Discussion open until November 1, 2007. Separate discussions
must be submitted for individual papers. To extend the closing date by
one month, a written request must be filed with the ASCE Managing
Editor. The manuscript for this paper was submitted for review and possible publication on February 18, 2006; approved on August 9, 2006. This
paper is part of the Journal of Architectural Engineering, Vol. 13, No.
2, June 1, 2007. ASCE, ISSN 1076-0431/2007/2-95104/$25.00.

buildings. The framework focuses on the early design stages and

covers ventilation with natural driving forces and mechanicalassisted ventilation. The goal is that through the development of
this framework designers can be better informed of the decisionmaking process for natural ventilation and a computer-based tool
is being developed to assist the process. The tool is planned to be
completed one year after the framework is ready. It is developed
on Windows platform with a user-friendly interface and will be
used in the early design phases.
Due to space limitations a full discussion of the structure is not
possible here although it will be available as a more complete
document. The overall structure and issues are presented.

Framework Overview
The decision-support framework is developed from three main
concerns: 1 the necessary inputs; 2 assessment procedures;
and 3 structure of the outputs. The inputs were determined by
identifying the critical issues of design when integration of natural ventilation is a goal. The critical issues were identified from
literature review and case studies of exemplars as well as interviews with architects and engineers with experience in design of
buildings for natural ventilation. The framework addresses both
all-natural and hybrid ventilation approaches, but focuses on the
decisions related to natural driving forces and asks in what conditions mechanical ventilation should be considered.
The assessment procedures link the inputs to the outputs and
provide the foundation for comparative analysis and recommendations from the framework. The assessment procedures were
also constructed from literature, case studies, and interviews.
The output from the framework must convey understandable
information for the intended user of the framework. For this, the
type of information and its format must be carefully considered.
Output may be categorized as binary yes or no, scalar low to
high or categorical. Which format is most appropriate for each
JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007 / 95

Fig. 1. Interactions between inputs and feasibility assessment, constraints and performance mandates. Note that a circle at each interaction means
the input is related to the intermittent result, constraint or the assessment. If the influence of an input on a performance mandate is through an
important intermediate result, no circle is shown at the intersection of the input and the performance mandate, but shown at both the intersection
between the input and the intermediate result, and between the intermediate result and the performance mandate.

output category must be decided. The framework begins with this

overall structure.

Critical Issues and Inputs to the Framework

Through review of related literature, case studies of naturally ventilated buildings, and interviews with experienced practitioners,
the critical decision-making issues and inputs were identified. The
critical issues include:
Architectural intention/integration;
Energy saving potential;
IAQ and health;
Thermal comfort;
Acoustics and background noise;
Safety and security; and
Cost.
These issues are identified from the major performance man96 / JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007

dates or constraints in the design of a naturally ventilated

building.
To address these issues, the inputs may be categorized as: 1
those associated with the owner and designers goals and objectives; 2 those associated with the design proposition; 3 those
associated with site and climate; and 4 the designers and owners relative levels of concern for each performance mandate, representing how the designer weighs the importance of each module
in a given project. The four categories condense the critical inputs
related to the design of natural ventilation. Examples of inputs
associated with each category are listed in Fig. 1.
Typically, for a given project, the performance mandates may
not have equal importance for decision making. Therefore the
relative level of concern for each of the performance mandates
energy, thermal comfort, IAQ, etc. must be input and used as
part of a decision-weighting process. These relative levels become normalized weights in the decision-making process as described later Table 2.

In addition to capturing the owner and architects design intentions as inputs, information concerning the site and climate are
needed. Among the site and climate data inputs are the temperature and humidity, prevailing wind speed and direction, solar
radiation, and noise and pollution sources interacted with site
conditions. If on-site measurements are available, they should
cover enough length of time to represent the actual environmental
conditions of the site, as limited on-site data may mislead the
assessments. Although on-site measurement is desirable, the
available data, most likely, are not the on-site measurements. To
be useful in the support framework, these inputs must often be
predicted, translated or adjusted for local effects. For example, the
following input processing may take place.
1. Wind at the building location: Wind velocity from meteorological stations available in typical meteorological year
archival files can be transformed and modified for microclimate effects for the terrain and height corrections at the
building location ASHRAE 2001. Although it would be
most desirable to determine these transformations using wind
tunnel studies or computational fluid dynamic CFD simulation, this is time, space, and/or cost intensive. Therefore for
the proposed decision-support framework, prediction models
for some typical scenarios are available. For example, the
algorithm calculating winds in the urban street canyons Ghiaus et al. 2006, and the approximate prediction of wind
speed reduction by trees Reed 1964 or other obstructions
are used. Corrections for wind direction may refer to the air
movement principles relating to typical topographies.
2. Temperature and humidity: Information concerning the ambient temperature is needed to assess potential energy savings, heat gain to the building and thermal comfort. In lieu of
on-site monitoring the TMY data should be adjusted for local
effects such as evaporative cooling by nearby bodies of
water, transpirative cooling by vegetation, or warming of the
air due to the heat island effect. Currently, no directly useful
methods are found for these adjustments. Research is needed
and is beginning in these areas. For example, the data of heat
island effect on temperature change in London was recently
measured Kolokotroni et al. 2006; Short et al. 2004; Watkins et al. 2002. Data such as this may be translatable to the
framework input processing procedures.
3. Noise intensity levels IL: Noise intensity levels at the
building site can be obtained from on-site measurement and
translated to noise contour maps. When this is not feasible,
some federal agencies provide standards and guidance to predict the exterior noise environment, such as highway traffic
noise model from the U.S. Department of Transportation, and
integrated noise model for aircraft noise from the Federal
Aviation Administration. The British standard method for
road traffic noise calculation DOT 1988 and geographic
information system tool integrated with a noise model as
well as local topography Keller 2006 are also sources that
are incorporated into the proposed framework. These sources
provide an approach to estimate the noise IL for the proposed
building site.
4. Outdoor air quality: Local outdoor air quality investigation is
required by ASHRAE Standard 62-2004. An observational
survey of the building site and its immediate surroundings
shall be conducted during hours the building is expected to
be normally occupied to identify local contaminants from
surrounding facilities that may be of concern if allowed to
enter the building. The input of outdoor air pollutants as the
results of the local outdoor air quality investigation will be

used to determine the viability of natural ventilation.

Algorithms estimating the amount of pollution caused by traffic or nearby emission are available. For example, pollution
downwind from a point source such as a stack, and a line source
such as a busy highway Wark et al. 1998, and pollution in a
street canyon within certain range of the height to width ratio can
be estimated from existing models Berlyand 1991.

Assessment
The decision-support inputs and outputs are related by a series of
both sequential and parallel assessment procedures. These assessment procedures may be Boolean or arithmetic with outcomes
that may be categorized as:
1. Binary yes or no. For example, feasibility assessment provides the result that natural ventilation is feasible or not;
2. Scalar. For example, performance assessment for energy savings potential can rate the estimated savings from low, medium or high, or in a numeric scale, such as 1 very low to
10 very high; and
3. Categorical. Such assessment provides categorized results,
such as the ranking of alternative natural ventilation approaches and categorized design suggestions.
The proposed decision-support framework has two primary
assessment algorithms. The first evaluates the feasibility of applying natural ventilation for the given inputs. If feasibility is
suggested then the relative performance of alternative natural
ventilation approaches is evaluated, ranked and compared, the
results of those assessment procedures are then presented as
recommendations.

Assessment Level 1: Feasibility

Modules assess the feasibility of natural ventilation for the given
inputs by analyzing the interaction of the adjusted critical site and
climate information with the design proposition Fig. 1. The algorithms in the modules are identified and developed from technical publications that are referred to the following sections. In
the proposed framework, the feasibility assessment consists of
four submodules:
1. Outdoor temperature and humidity assessment;
2. Wind assessment;
3. Site noise assessment; and
4. Pollution assessment.
If the outcome from any of the four modules concludes that
natural ventilation only is not feasible, no further assessment is
made. Certain remedial solutions may be suggested, including
hybrid and mechanical ventilation.
Processing within the modules is as follows.
Outdoor Temperature and Humidity
The outdoor air temperature and humidity can be used to assess
potential energy savings and estimate thermal comfort conditions.
Similar to the procedures used in climate analysis tools such as
Climate Consultant, climate data can be graphically analyzed by
overlaying hourly climate data on a psychrometric chart with the
thermal comfort zone identified. The process of analysis by overlay is shown in Fig. 2 with Climate Consultant Department of
Architecture 2006.
In this module, based on the methods by Emmerich et al.
JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007 / 97

Fig. 2. Climate consultant result for TMY weather data overlaid with psychrometric thermal comfort zone

2001, first, the buildings cooling balance point temperature and

the heating balance temperature are calculated based on the
boundary of comfort zone of ASHRAE Standard 55 2004, with
only infiltration considered. To avoid concerns for condensation
and microbiological growth, the humidity upper limit for natural
ventilation is set to 70% relative humidity or 0.012 humidity
ratio, whichever is higher Avens and Baughman 1996; Clarke
et al. 1999. The range defined by the two balance point
temperatures and the humidity limit are superimposed on the psychrometric chart. Then the hourly conditions for temperature and
humidity from the TMY file are overlaid. When outdoor temperature falls in the range between the two balance point temperatures, no ventilative cooling is necessary, but ventilation is needed
to maintain IAQ. If the outdoor temperature is above the cooling
balance temperature but humidity is below the limit, ventilative
cooling is needed. Then the upper temperature limit for natural
ventilation is based on the allowed indoor air speed and adjusted
by the building heat gain conditions and estimated ventilation air
flow rate as determined by the wind assessment module. Thus a
range of outdoor conditions can be identified when natural ventilation is feasible. A comparison of the annual number of hours
falling within this range to a minimum number of hours established by the designer can be used to determine the feasibility of
achieving performance benefits from natural ventilation. As natu98 / JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007

ral ventilation can be a cooling system and a ventilation system, it

is useful even in cold seasons to maintain IAQ. However, in this
module, the hours below heating balance temperature are not
counted. For those hours when the ventilative cooling is inadequate, the feasibility of night ventilation with thermal mass coupling should be assessed.
The adaptive model for thermal comfort was not selected for
use in the feasibility assessment because 1 the adaptive model
exhibits the relation of indoor comfort operative temperature
ranges to the outdoor temperature, the former is difficult to determine with inadequate design information at the feasibility assessment stage and 2 as the adaptive model assumes the occupants
can adapt to the local environment by opening windows, changing clothes, etc., some occupancy types and building locations do
not allow such adaptation. The adaptive model is used later in the
performance assessment module.
Wind
The feasibility of natural ventilation should also be assessed in
terms of wind speed and direction with respect to the microclimate, building geometry and orientation, and size and placement
of openings. After adjusting for topography and microclimate
conditions, the module analyzes the pressure distribution on the

Table 1. Natural Ventilation Strategies, Ventilation Components in the Flow Path

Flow path
Supply natural/mechanical

Strategies
Single-sided ventilation
cross flow only stack
ventilation only cross
ventilation with stack
effect
Cross ventilation with
wind chimney and stack
effect

Inlets
Windows, wind scoops
or vents, with manual/
automatic control

Exhaust natural/mechanical

Possibly through
ventilation components
and building elements

Possibly through
ventilation components
and
building elements

Vertically: atria, shafts

Vertically: atria, stair

cases, shafts, wind tower,
double-skin facade

Horizontally: double-skin
facades, ducts, corridors,
underneath floor supply
plenum

Horizontally: corridors,
ducts

building surfaces caused by the prevailing wind. Ideally, this

would be accomplished by wind tunnel tests or CFD simulation,
but these approaches are typically not appropriate in the early
stages of design. As an alternative, for buildings with approximate
rectangular shapes, published data of pressure distribution on
building surfaces for varying wind speed and direction are available ASHRAE 2001; Santamouris and Asimakopoulos 1996 For
buildings with irregular shapes, adjustments must be made. Exactly how to adjust is not available in the literature and needs
further study. If the average wind speed is too low, a stack driven
or a hybrid system will be suggested.
Noise
In the noise assessment module, there are algorithms estimating
the interior noise intensity levels with input of the outdoor noise
intensity level and the proposed envelope characteristics and ventilation openings and windows. The noise feasibility module proceeds with:
1. Estimate the total transmission loss of a composite barrier
Sharland 1972;
2. Calculate the noise reduction between two spaces separated
by the barrier Stain and Reynolds 1992; and
3. With the required indoor NC levels by building types, estimate the required sound reduction and compare with the calculated noise reduction in Step 2.
Ideally, noise intensity for different octave bands at the building site should be known to determine the acceptable STC rating.
If noise levels appear to be too high then some remedial design
suggestions are provided, e.g., relocation the openings, doubleenvelope, outside barrier, acoustical louvers and dampers with
automatic control and intermittent ventilation, or active noise control.
Pollution
Table 41 of ASHRAE Standard 62-2004 is used to assess if the
outdoor air can be used for natural ventilation. The regional outdoor air quality compliance status for the United States as found
in the air quality system Database from the U.S. Environmental
Protection Agency USEPA website is useful for determining the
ambient air quality USEPA 2006. Local pollutant sources and

Outlets
Windows, vents, with
manual/automatic control

their location relative to the building and prevailing winds should

be assessed on-site. The results of local outdoor air quality investigation are compared with Table 41.
Building spaces typically have indoor pollutant sources and
consequently potentially high indoor pollutant concentrations. For
natural ventilation this has at least three potential consequences.
First and easiest to address would be to not include these spaces
in the natural ventilation solution and provide mechanical ventilation. However, this would seem to defeat the intention of natural
ventilation. The premise here is, if natural ventilation is properly
designed, the adequate flow rates can be achieved for relatively
clean or polluted spaces. Second, if applied, the natural ventilation solution must provide both adequate flow volume and ventilation effectiveness to dilute and remove the pollutant. Third, the
air flow path must be considered to prevent cross-contamination.
For this, compartmentalization and zoning may be an effective
strategy.
For the decision-support framework estimation of IAQ is
made by estimating the indoor air flow rate.

Assessment Level 2: Comparison of Alternative

Natural Ventilation Approaches
If the result from any one of these four feasibility assessment
modules suggests that natural ventilation is not viable, then
redesign strategies are suggested, but detailed performance assessment is not carried out. If, however, natural ventilation is
determined to be feasible, then assessment proceeds to a more
detailed comparative analysis.
Natural ventilation strategies may be categorized as winddriven ventilation or buoyancy-driven stack effect ventilation.
Strategies may be further subcategorized as single-sided, cross
flow only, stack ventilation only, cross ventilation with stack effect, and cross ventilation with wind chimney and stack effect, as
shown in Table 1. Table 1 summarizes the natural ventilation
strategies, the driving forces, possible flow path that includes inlets, ventilation components and building elements, and outlets.
Which approach is most appropriate for a given set of inputs is a
concern for the decision-making process. Comparison of these
alternatives should consider the inputs such as building geometry
and prevailing wind speed and direction. Therefore the decisionJOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007 / 99

Table 2. The Performance Elements Pij, a Set of Normalized Weightings, and the Overall Performance Scores of Two Alternatives

Weightings j
Alternative 1
Alternative 2

Energy
saving

Thermal
comfort

IAQ

Acoustic
performance

Cost

Total

0.3
P11 = 0.7
P21 = 0.6

0.2
P12 = 0.65
P22 = 0.8

0.2
P13 = 0.9
P23 = 0.9

0.1
P14 = 0.5
P24 = 0.75

0.2
P15 = 0.85
P25 = 0.4

1.0
P1 = 0.74
P2 = 0.675

support framework should allow for comparison of these alternatives for the given inputs and constraints by evaluating the performance mandates.
Application of Constraint Functions
All of the possible natural ventilation approaches and system designs are assessed by first applying constraint functions to each of
the alternatives. The scope of the constraints in the framework is
introduced as follows.
Architects and Clients Preference on Natural Ventilation
Options
The architects and the clients preferences for the level of expression of natural ventilation systems and their preferences for natural ventilation options determine the choice of natural ventilation
elements to be integrated.
Safety and Security
The risk that fire may spread along natural ventilation paths is a
concern. Fire codes can regulate a natural ventilation system design with requirements on the opening size and locations in exterior walls. The need for compartmentalization may also limit the
application of natural ventilation. However, the fire code may not
necessarily be a major barrier to the implementation of natural
ventilation. Strategies such as fire-warning systems, and smoke
alarm systems can detect fire and signal to the occupants, whereas
openings can be automatically or manually controlled to close, or
open for smoke ventilation. For example, ventilation openings
penetrating fire areas can be designed with automatic-closing fire
doors, shutters, or dampers. In some cases, fire-suppression systems may be necessary.
Opening size and location should prevent injury from falling
by proportioning the length and width of openings such that an
adult or child cannot pass through. This would be especially important in elementary schools and nurseries.
Security standards vary with building type and may limit natural ventilation from certain types of buildings or certain building
spaces DoD 2003, 2005; GSA 20053, For example, operable
windows are usually restricted in the public area of courthouses
Nadel 2004. In federal buildings, openings tend to be minimized
and protected as a security measure. In the San Francisco federal
building, security concerns mandate that the lower, more vulnerable floors be completely sealed. McConahey et al. 2002 The
levels of protection depend on the building type, acceptable levels
of risk, and decisions made based on recommendations from a
project-specific threat and vulnerability assessment, and risk
analysis. ASHRAE provides a general risk management guide for
health, safety, and environmental security under extraordinary
incidents for buildings ASHRAE 2003. For the design of ventilation systems related to security, if pressurization, isolating bigthreat areas and protection from external and internal release of
chemical, biological and radioactive agents are necessary according to risk and vulnerability assessment, a hybrid ventilation sys-

tem with integrated building automation and control system that

enables monitoring and control of traditionally standalone building systems is appropriate, because isolation zones are required
and air filtration and pressurization are needed in emergencies.
For most nonresidential buildings such as commercial office
buildings, carefully designed openings can overcome the security
problem.
Zoning Requirement
The zoning requirements that potentially influence the design of
naturally ventilated buildings include: setbacks, lot coverage,
height restrictions, building area and development densities. For
example, the density requirements may influence the available
winds; the height restrictions may limit the use of ventilation
stacks.
Additionally, size and proxemic relationships among functional zones may potentially limit the implementation of natural
ventilation by regulating the geometry of the building. For example, the need to place large zones adjacent to each other may
increase the overall depth of the building plan making it difficult
to develop a ventilation flow path.
Air Flow Reduction with Insect Screens and Other Devices
If used at ventilation openings, insect screening may cause significant pressure loss, e.g., rough screens can reduce air flow
more than 60%, and smooth screens can reduce the air flow 35%
on average Bowen et al. 1981. Some commercial CFD model
allows the screen to be modeled with free area ratio. Sometimes it
is not practical to compensate for this loss by increasing the opening areas, thus mechanical assistance may be needed. Similarly, to
reduce noise transmission, and provide sound buffering, building
components such as double envelopes, shafts, plenums, atria,
acoustic louvers and acoustical attachment, and insulation may
induce air flow resistance. The device for shading or glare control
can also add flow resistance. CFD may be necessary to study the
degree of flow reduction by these components or devices.
Weather
Another issue that potentially constrains the application of natural
ventilation is weather, such as monsoon rain in tropical climates
and sand storms in hot arid regions.
Performance Mandates
The comparison of alternative approaches to natural ventilation
continues by assessing the critical performance mandates, or the
degree of achievement of required goals. The mandates in the
framework include:
Energy savings;
Thermal comfort;
Acoustic control;
Indoor air quality; and
Cost.
These assessments use hourly TMY data and mean daily if

100 / JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007

available or monthly wind speed and direction. Estimating indoor air flow rate is based on the Chartered Institution of Building
Services Engineers method 1997 or multi-zone network models
Emmerich and Dols 2003; Feustel 1998, by using the pressure
distribution from wind and assumed values for resistance and
pressure loss through the space including 1 the loss at openings
in the form of discharge coefficients based on experimental data
compared with those borrowed from the data traditionally used
for fluid flow in pipes Favarolo and Manz 2005; Flourentzou
et al. 1998; Heiselberg et al. 2001; 2 the loss in shafts and ducts
ASHRAE 2001; and 3 the loss at other components such as
screens.
Occupant health and productivity are other important issues
that can be considered as performance mandates since naturally
ventilated buildings can potentially achieve better indoor air quality, whereas window openings also provide daylight and views
that improve occupants health and a biophilia connection to outdoors. However, occupant health and productivity are difficult to
estimate and quantify especially in the early design stage. They
are comprehensive metrics resulting from various factors including thermal comfort, acoustical environment, IAQ, view, etc.
Therefore, to simplify the problem, users may add more such
factors that are concerned with occupant health and productivity
in the early design stage.
Energy Savings
A potential selling feature of natural ventilation is lower energy
consumption. The decision-support framework proposes a simplified assessment approach, good for comparative assessment rather
than prediction of absolute savings of each alternative. The steps
include:
Estimate the pressure distribution at the building envelope;
Estimate the air flow rate through interior spaces;
Estimate the heat removal rate with natural ventilation versus
the rate of heat gain with maximum allowed indoor temperature and outdoor temperature;
Sum the number of hours per year when natural ventilation
can effectively remove heat from the building;
Estimate the heat removal rate for each applicable hour; and
Sum and compare the heat removal for all applicable hours of
all alternative approaches.
If hybrid ventilation or other cooling strategies are used together with natural ventilation, a whole building simulation with
thermal-air flow models is needed.
Thermal Comfort
Performance assessment of thermal comfort is constrained by the
conditions specified in ASHRAE Standard 55 2004. The combined effects of temperature, humidity, radiation and air movement are assessed. The adaptive model Brager and Dear 2000 is
used to assess indoor operative temperatures. As suggested by
ASHRAE Standard 55 2004, for more extreme conditions, the
conditions for draft and allowable limits for turbulence intensity
can be determined according to the requirement on the local discomfort criteria. For comparison of alternative natural ventilation
strategies, the constraints can be applied to a proposed assessment
procedure that might include the following steps using TMY data
and the given inputs:
Estimate the pressure distribution at the building envelope;
Estimate the air flow rate through the building;
Estimate the heat removal rate versus the rate of heat gain;
Estimate the temperature rise in the occupied zone;
Estimate the humidity level currently no simplified methods

for estimating the indoor humidity level are available. For the
early design phase, in this framework, if the outdoor relative
humidity is compliant to the requirement in the feasibility assessment for the indoor humidity level, it is considered acceptable, except for buildings with high internal latent heat gain,
such as assembly buildings with high occupant density; in this
case, CFD simulations may be needed;
Compare the resulting air flow rates, temperature, humidity
and the risk of draft to the constraint functions; and
Sum the annual number of hours when the constraint functions
are satisfied.
Indoor Air Quality
A third performance mandate is indoor air quality. If the proposed
building passes the preliminary IAQ feasibility assessment, then
the relative performance of the alternative natural ventilation
strategies should be evaluated. The IAQ assessment includes
evaluation of two performance criteria: ventilation flow rate and
ventilation effectiveness from ASHRAE Standard 62 2001. For
ventilation flow rate the following steps are proposed:
From hourly TMY data for wind speed and direction, estimate
the pressure distribution at the building envelope;
For the descriptive input for approximate size and location of
inlets and outlets estimate the air flow rate through the space;
Estimate the average flow rate per square foot of floor area and
volumetric air change per hour; and
Compare to ASHRAE Standard 62 minimum values Table 2,
ASHRAE Standard 62-2001
Addressing ventilation effectiveness concerns with relative
size, uniformity of distribution, and location of inlets and outlets.
For the proposed decision-support framework, ventilation effectiveness is assessed and the output includes recommendations to
insure good mixing or to use displacement ventilation. It is acknowledged that wind is typically not a criterion for selecting
TMY data. However, until another data source is identified the
TMY files will be used as a compromise.
Acoustic Performance
To assess noise levels inside the building, estimated indoor noise
intensity levels are compared to published noise criteria ratings
for the given building functions. The suggested procedure estimates the attenuation of the ambient noise levels through the
envelope openings into the occupied zone. Then for the worstcase condition i.e., street level usually the estimated noise levels
are compared to the constraint functions. Sounds with information
content transmitted from major indoor sources such as an adjacent
meeting room through ventilation openings is also assessed for
speech privacy.
Cost
Ideally a detailed life-cycle cost evaluation of alternative natural
ventilation system approaches should be performed. In early design phases, relative comparison of design features and natural
ventilation system elements should be assessed. In the framework, because the comparison is only between those natural or
hybrid ventilation approaches, the assessment can be simplified
based on the method of Ghiaus and Roche 2005, in which only
the incremental cost and benefits related to these alternatives are
considered. The major inputs of cost comparison include:
1. Initial cost
a.
The cost of ventilation systems and building elements
associated with natural ventilation, which includes operable windows, vents, chimneys, transoms, screens,

JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007 / 101

louvers, etc. These also include the cost increase associated with building components such as an atrium, if it
is used for natural ventilation.
b. Devices such as monitoring and control system if any.
c.
Mechanical assistance system if any;
2. Operation cost
a.
Maintenance cost.
b. Energy consumption such as the energy used for mechanical assistance.
c.
Energy saving;
3. Replacement cost: The replacement cost of some components with short lifetimes should also be considered. To simplify the method, the replacement cost of components with
long lifetimes may be ignored;
4. Discount rate; and
5. Time period investigated.
The calculation can be easily done with a spreadsheet. The
discounted cost or benefit is calculated for each year and then the
total discounted cost or return is summed for the total cost or
returns for the period investigated.
A major hidden cost or benefit is the cost resulting from the
impact on worker health and productivity. Health and productivity
are influenced by thermal comfort, IAQ and acoustics. However,
productivity can be difficult to quantify. The quantitative relations
between productivity and the performance mandates are currently
not available. The default assessment excludes productivity and
focuses on the life-cycle cost of building system itself. If considering the cost of employment, a linear relation of productivity to
the performance mandates can be assumed for the preliminary
assessment. Thus the cost of total salary adjusted by productivity
can be estimated by multiplying a coefficient that is inverse to the
standardized value of the sum of the three performance scores.
The alternative with the best overall performance will tend to be
the one with the higher performance of thermal comfort, indoor
air quality and acoustics, when compared to the results of the
default assessment with productivity excluded.

Fig. 3. Vectors of the performance elements Pij

weightings of various typical uses of buildings from experienced

professionals as benchmark and default values of weightings in
the framework.
Ranking
Because ranking design approaches needs to examine several performance factors, the multipurpose optimization method is used
to assess the performance of alternative natural ventilation system
design approaches by weighting these factors. For alternative i
i = 1 , . . . , i and factor j j = 1 , . . . , j, the performance of alternative i on factor j is defined by Pij.
For example, the vectors of the performance elements Pij are
illustrated with a radar diagram Fig. 3. With a set of normalized
weightings j kj=1 j = 1, the overall performance score of an
alternative i is:
j

Pi =

1 jpij

The example is shown in Table 2. Thus, system design approach

alternatives can be ranked.

Weighting
Before comparisons between alternatives can be made, two additional decision making steps are necessary. These include first
obtaining a normalization scale to each performance mandate outcome. For example, estimation of thermal comfort conditions or
ventilation air flow rate can be scaled from 0 to 1, 0 being total
failure to meet a performance mandate and 1 being total compliance. Or, a set of performance outcomes can be standardized as:
fx j = x j/max x j

fx j = x j max x j/max x j min x j

or

where max x j and min x j indicate the maximum and minimum

value observed for objective x j among all alternatives Nijkamp
et al. 1990. In this standardization method, only comparison between the values of fx j is important to the evaluation.
The second step involves weighting the relative importance of
each performance mandate. For example, when comparing and
selecting between natural ventilation alternatives, if the building
is a library, the acoustic performance may have higher relative
importance to the other mandates. Although these weightings are
often applied intuitively, surveys are planned to gather these

Output
The users of the framework are architects and engineers involved
in the process of natural ventilation design. The usefulness of
assessment algorithms and the levels of detail of the outputs must
be appropriate for the early phases of design.
Each decision-making module includes input, process algorithms and output. The output can be feasibility assessment,
ranking of alternative, natural ventilation approaches and design
suggestions. Suggestions include the design and sizing recommendation for the natural ventilation system features, and control
strategies. The final output of the framework is a ranking of
alternatives.
The general structure and process of the decision-support
framework is illustrated in Fig. 4.
Fig. 1 summarizes the interactions between inputs and issues
of feasibility assessment, constraints and performance assessment. It demonstrates the internal relation of those issues. In this
framework, the interaction and the tradeoff between the performance mandates are analyzed qualitatively by tracing their common related inputs or intermediate results in Fig. 1. The results
are given as design suggestions. The interactions can be quantitatively studied in the future.

102 / JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007

Fig. 4. General structure and process of the decision-support framework

Conclusion and Summary

For natural ventilation in nonresidential buildings, the challenge
for architects and engineers is that collaborative decisions must
begin in the early design phase. This demands a decision-support
framework to assist the process, integrating both architectural and
engineering concerns. In such a framework, since comparison between alternatives is done to find the best for the early stages of
design, relative performance is more of a concern than absolute.
This can also make the assessment process simpler. In fact, absolute performance prediction is unrealistic in early design stages.
The framework is being adapted to a computer-based tool
based on the modular design. The framework is a decision support system intended to give the designer useful feedback for their
proposed design. The user can change the design according to the

outputs of the framework. Therefore, the design process with the

support of the framework is iterative. Through the iterative process, users can improve their design when natural ventilation is a
goal.
Note: The writers invite feedback and comments on the
framework through the writerss contact information.

References
American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. ASHRAE. 2001. Air flow around buildings. ASHRAE
handbook, fundamentals, Atlanta, 16.116.11.
American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. ASHRAE 2003. Rep. of Presidential Ad Hoc Commit-

JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007 / 103

tee for Building Health and Safety under Extraordinary Incidents on

Risk Management Guidance for Health, Safety, and Environmental
Security under Extraordinary Incidents, Atlanta.
Avens, E. A., and Baughman, A. V. 1996. Indoor humidity and human
health. Part I: Literature review of health effects of humidity influenced indoor pollutants. ASHRAE Trans., 1996, 3951.
Berlyand, M. E. 1991. Numerical methods of air pollution forecasting. Prediction and regulation of air pollution, Kluwer Academic,
Dordrecht, The Netherlands, 9194.
Bowen, A., Fairey, P. W., and Chandra, S. 1981. Passive cooling by
natural ventilation, world literature review and annotated bibliography, Florida Solar Energy Center FSEC, Orlando, Fla.
Brager, G. S., and Dear, R. J. D. 2000. A standard for natural ventilation. ASHRAE J., 4210, 2128.
Chartered Institution of Building Services Engineers CIBSE. 1997.
CIBSE applications manual AM10: Natural ventilation in nondomestic buildings, London.
Chen, Q. 2004. Using computational tools to factor wind into architectural environment design. Energy Build., 3612, 11971209.
Clarke, J. A., Johnstone, C. M., Kelly, N. J., McLean, R. C., Anderson, J.
A., Rowan, N. J., and Smith, J. E. 1999. A technique for the prediction of the conditions leading to mould growth in buildings. Build.
Environ., 23, 404410.
Department of Architecture and Urban Design. 2006. Energy design
toolsClimate consultant 3. Univ. of California, Los Angeles,
http://www2.aud.ucla.edu/energy-design-tools/ July, 2006.
Department of Defense DoD. 2003. Unified facilities criteria (UFC),
DoD minimum anti-terrorism standards for buildings, Washington,
D.C.
Department of Defense DoD. 2005. Unified facilities criteria (UFC),
security engineering: Procedures for designing airborne chemical,
biological, and radiological protection for buildings, Washington,
D.C.
Department of Transportation DoT. 1988. Calculation of road traffic
noise, London.
Emmerich, S. J., and Dols W. S., 2003. LoopDA: A natural ventilation
system design and analysis tool. 8th Int. IBPSA Conf., Eindhoven,
The Netherlands.
Emmerich, S. J., Dols, W. S., and Axley, J. W. 2001. Natural ventilation
review and plan for design and analysis tools, National Institute of
Standards and Technology, Washington, D.C.
Favarolo, P. A., and Manz, H. 2005. Temperature-driven single-sided
ventilation through a large rectangular opening. Build. Environ.,
405, 689699.
Feustel, H. E. 1998. COMISAn international multizone air-flow and
contaminant transport model, Lawrence Berkeley National Laboratory LBNL, Berkeley, Calif.
Flourentzou, F., Maas, J. V. D., and Roulet., C. A. 1998. Natural ventilation for passive cooling: Measurement of discharge coefficients.
Energy Build., 27, 283292.
General Services Administration GSA. 2005. Facility standards for
the public building services.

Ghiaus, C., Allard, F., Santamouris, M., Georgakis, C., and Nicol, F.
2006. Urban environment influence on natural ventilation potential. Build. Environ., 414, 395406.
Ghiaus, C., and Roche, L. 2005. Whole life costing of ventilation
options. Natural ventilation in the urban environment, assessment
and design, Earthscan, 227235.
Heiselberg, P., Svidt, K., and Nielsen, P. V. 2001. Characteristics of air
flow from open windows. Build. Environ., 367, 859869.
Jones, J., and West, A. W. 2001. Natural ventilation and collaborative
design. ASHRAE J., 4311, 4651.
Keller, K. 2006. Simple GIS tools for traffic and transit noise studies.
http://gis.esri.com/library/userconf/proc02/pap0505/p0505.htm.
Aug. 11, 2006
Kolokotroni, M., Giannitsaris, I., and Watkins, R. 2006. The effect of
the London urban heat island on building summer cooling demand
and night ventilation strategies. Sol. Energy, 804, 383392.
McConahey, E., Haves, P., and Christ, T. 2002. The integration of
engineering and architecture: A perspective on natural ventilation for
the new San Francisco federal building. ACEEE 2002 Summer Study
on Energy Efficiency in Buildings, Grove, Califa.
Nadel, B. A. 2004. Courthouse security. Building security, handbook
for architectural planning and design, McGraw-Hill, New York, 6.1
6.24.
Nijkamp, P., Rietveld, P., and Voogd, H. 1990 Assessment of priority
weights and preferences. Multicriteria evaluation in physical planning, Elsevier Science, Amsterdam, The Netherlands, 4049.
Reed, R. A. 1964. Tree windbreaks for the central great plains. Agriculture handbook #250, U.S. Department of Agriculture, Washinghton, D.C.
Santamouris, M., and Asimakopoulos, D., eds. 1996. Natural ventilation: Appendix D. Passive cooling of buildings, James & James Science Publishers, London, 301306.
Sharland, I. 1972. The calculation of noise levels. Woods practical
guide to noise control, Woods of Colchester Ltd., Colchester, U.K.,
7177.
Short, C. A., Lomas, K. J., and Woods, A. 2004. Design strategy for
low-energy ventilation and cooling within an urban heat island.
Build. Res. Inf., 323, 187206.
Stain, B., and Reynolds, J. S. 1992. Building noise control. Mechanical and electrical equipment for buildings, Wiley, New York, 1365
1376.
U.S. Environmental Protection Agency USEPA. 2006. Air quality
system AQS database. http://www.epa.gov/air/data/aqsdb.html
Aug. 11, 2006.
Wark, K., Warner, C. F., and Davis, W. T. 1998. Dispersion of pollutants in the atmosphere. Air pollution, its origin and control, Addison
Wesley, Menlo Park, Calif., 146161, 171173.
Watkins, R., Palmer, J., Kolokotroni, M., and Littlefair, P. 2002. The
London heat island: Results from summertime monitoring. Build.
Services Eng. Res. Technol., 232, 97106.

104 / JOURNAL OF ARCHITECTURAL ENGINEERING ASCE / JUNE 2007